> ## Documentation Index
> Fetch the complete documentation index at: https://docs.openp.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Cherry Studio

> Add OpenPAI as a model provider in Cherry Studio

[Cherry Studio](https://www.cherry-ai.com/) is a cross-platform, multi-model AI desktop client.

## Configuration steps

<Steps>
  <Step title="Open settings">
    Gear icon at the bottom-left → **Model services**.
  </Step>

  <Step title="Choose an OpenAI-compatible provider">
    In the provider list, choose **OpenAI** or add a new **custom OpenAI-compatible** provider.
    We suggest naming it `OpenPAI`.
  </Step>

  <Step title="Fill in connection details">
    * **API address / Base URL**: `https://openp.ai/v1`
    * **API key**: paste the `sk-` key from the OpenPAI console
    * **Models**: click **Add model** and enter the model IDs you want (e.g. `gpt-5.5`, `claude-opus-4-8`, `gemini-3.1-pro-preview`)
  </Step>

  <Step title="Click check connection / enable">
    Cherry Studio makes a `/v1/models` call to verify the key works.
  </Step>

  <Step title="Start a new chat">
    Back on the main screen → new chat → pick the OpenPAI model you just configured and start using it.
  </Step>
</Steps>

## Tips

* Cherry Studio supports **enabling multiple providers at once** — you can make OpenPAI the primary service and use others as fallback.
* To use the Anthropic native protocol (`/v1/messages`) for advanced features like thinking and cache\_control, choose an **Anthropic-compatible** provider in Cherry Studio and set the Base URL to `https://openp.ai` (without `/v1`).
* The Gemini native protocol is similar: choose the **Gemini** provider and set the Base URL to `https://openp.ai`.

## FAQ

<AccordionGroup>
  <Accordion title="It says 401 invalid_api_key">
    * Check the key was pasted in full (no trailing spaces / newlines).
    * Confirm in the OpenPAI console that the token isn't disabled / expired.
  </Accordion>

  <Accordion title="The model list is empty">
    * Click the **Fetch models from provider** button; Cherry Studio pulls `/v1/models`.
    * If it's still empty, manually **add a model** and enter the model ID.
  </Accordion>

  <Accordion title="How do I tell OpenPAI apart from official OpenAI?">
    Distinguish them by provider display name (e.g. `OpenPAI` vs `OpenAI`).
    The chat history at the bottom-left shows the provider name used.
  </Accordion>
</AccordionGroup>